NO. I SCALES OF AFRICAN FISHES — COCKERELL 5 



(2) Rather large white scales, with the apical circuli forming conspicuous 



strong ridges ; the basal circuli, as usual, very fine. 

 Alestes aiHnis Giinther. Wabbi System (O. Neumann). 



Only one basal radius ; apical region with the radii branching and anasto- 

 mosing quite irregularly, all very strong; scale about 7 mm. long and 

 8 broad. 

 Alestes imheri Peters. Lake Nyassa (Capt. E. L. Rhoadcs). 



Thinner scales than those of A. affinis ; basal radii three; apical region 

 quite as in /i. aMnis. Scale about as broad as long. 



(3) Rather small (6 to 7 mm. diameter) whitish scales of a rather broad 



type, having the outline of a little more than a half circle; apical 

 sculpture feeble ; basal radii three or four. There is no sign of the 

 apical structure described under A. afUnis. These scales are certainly 

 nearer to the macrophthahniis group than to the aMnis group. 



Alestes (BracJiyalestes) longipinnis Steindachner. Sette Cama, Gaboon 

 (Gerrard). 



Alestes chaperi Sauvage. Kotchwah River {R. B. N. Walker). 



These two were united by Giinther, and certainly there is no difference in 

 the scales. 



(4) Small scales about 6 mm. diameter or less, without very marked 



characters. 

 Alestes sadlcri Boulenger. Entebbe, Lake Victoria (Degen). 



Polygonal areas present or absent; apical circuli rather strong. 

 Alestes lateralis Boulenger. Kafue River, Upper Zambesi (T. Codrington). 



Scale small, but not from a fully-grown fish. Only one basal radius. 

 Alestes intermedius Boulenger. Kribi River, Cameroon (G. L. Bates). 



Scale rather broad, with five basal radii; apical sculpture very feeble; 

 lateral circuli, especially in the subdiscal region, mmch less dense 

 than usual, and minutely wavy or zigzag. 

 Alestes nurse Riippel. River Nile. 



Basal radii variable up to about six; pob^gonal areas present or absent. 

 Alestes haremose Joannis. River Nile. 



Basal radii one or two; polygonal areas not observed; scale in general 

 similar to that of A. nurse, but differing in important details, as 

 shown below. 

 The scales described above were all taken from the vicinity of the lateral 

 line, about the level of the beginning of the dorsal fin. The following notes 

 describe the variation observed on the different parts of the fish. 

 Alestes macrolepidotus. This species is remarkable for the very large scales, 

 and the posterior situation of the first dorsal fin. The borders of the 

 scales are beautifully purplish-iridescent, but the fish is not silvery 

 like A. nurse and haremose. The coarse apical circuli become 

 entirely longitudinal, suggesting the corresponding structures in 

 Amia calva, in which the progenitors of the circuli are longitudinal 

 strands. The apical radii largely anastomose transversely, resulting 

 in the formation of very irregular transverse markings, few in num- 

 ber. In the ventral region the sculpture is more profuse, and the 

 apical field develops a sort of cancellation, in which the longitudinal 

 lines appear to represent modified circuli, and the transverse modi- 

 fied radii. The anterior dorsal scales are again quite different, with 



